Stability and Repeatability Analysis of a Phase-Modulated Optical Fibre Sensor for Transformer Oil Ageing Detection.

Transformer oil ageing alters key physicochemical properties, notably the refractive index (RI), due to physical, particulate, and chemical changes. As a result, refractometric fibre-optic sensors have gained attention for enabling real-time monitoring and overcoming the limitations of traditional offline diagnostics. This study explores the use of a Fabry-Pérot phase-modulated fibre optic sensor (FISO FRI RI Sensor) for in-situ ageing detection in four industrial transformer oils: natural ester, synthetic ester, Nytro Bio 300X (vegetable-based), and Polaris GX (mineral-based). The oils were thermally aged under controlled conditions following degassing and drying. The sensor performance was evaluated using key metrics, including repeatability, thermal response, settling time, and linearity. Results show high repeatability (with standard deviations below 7 × 10-5 RIU and repeatability coefficients under 2 × 10-4 RIU), stable thermal response (~0.0004 RIU/°C), and strong thermal linearity (R2 > 0.99) across all samples. Natural ester and Nytro Bio 300X exhibited the most stable and consistent sensor responses, while synthetic ester and mineral oils showed greater variability due to temperature-induced RI shifts. These findings demonstrate the reliability and precision of this Fabry-Pérot phase-modulated sensor for online transformer oil condition monitoring, with strong potential for integration into smart grid diagnostics.